This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Borderline personality disorder (BPD) is a persistent and severe mental disorder with high mortality, morbidity and levels of distress. Emotional dysregulation or reactivity to environmental events, particularly events such as separations, frustrations, or losses is considered a defining feature of this disorder. This disorder has received little investigative scrutiny and the neural correlates of abnormal emotional processing are not understood. Startle eyeblink modification (SEM) methodology provides a non-verbal, quantitative tool for the study of emotional processing. When a brief startle stimulus is presented during an unpleasant stimulus, the amplitude of the startle response is potentiated;a pleasant stimulus occasions a relative inhibition of the startle response;and a neutral stimulus is intermediate. This study uses SEM and functional neuroimaging (event-related fMRI and FDG-PET) to explore the intensity, time course, habituation and functional neuroanatomy of emotional responses in three rigorously screened groups: 26 BPD patients (with no comorbid Axis I disorder or Cluster A or C personality disorders), 26 non-BPD patients (with a diagnosis of schizotypal personality disorder and no comorbid Cluster B or C personality disorders or Axis I disorder) and 26 healthy controls. During both the fMRI scan and FDG-uptake for PET, participants will view an intermixed series of unpleasant (e.g., hate), pleasant (e.g., happy) and neutral (e.g., view) words. SEM amplitude difference scores (e.g., unpleasant-pleasant) measured during simultaneous FDG uptake will provide a measure of the intensity, time course, and habituation of emotional processing of unpleasant stimuli and be correlated with regional glucose metabolism. fMRI BOLD response difference scores in key brain regions will provide measurement of the intensity and time course of brain activation. We focus on three brain regions thought to be involved in emotional processing: amygdala, anterior cingulate, and prefrontal cortex (orbitofrontal and dorsolateral). These regions are part of a brain circuit thought to mediate emotional experience and play a significant role in the expression of BPD symptoms. This proposed combination of a psychophysiological and functional neuroanatomy approach holds the potential to provide much-needed information on normative aspects of emotional processing and insights into the neuroanatomy underlying the emotional dysregulation observed in BPD. Hypothesis: Hypothesis (1a): Greater intensity of the startle response magnitude during the unpleasant words in BPD patients than normal controls, but there will be no between-group differences for the pleasant or neutral stimuli. Test: MANOVA on SEM percent change scores will be conducted with Group as the between-group factor and Word and Probe condition as repeated measures: Group (BPD patients vs. Controls)[unreadable]Word condition (unpleasant, pleasant, neutral)[unreadable]Probe position (3000ms, 4000ms, 5000ms). We predict a significant Group[unreadable]Word condition interaction. Next, Fisher LSD tests will be used post-hoc to determine whether our hypothesis of BPD>N SEM during unpleasant words, but not pleasant or neutral is observed. Hypothesis (1b): BPD patients will exhibit greater affective potentation of startle during unpleasant words at the later probe position and (1c) this potentiation in BPD patients will show slow habituation. Test for 1b: The MANOVA in 1a will test this hypothesis and we predict a significant Group[unreadable]Word condition[unreadable]Probe position interaction. Post-hoc tests showing BPD>N SEM at 5000 ms probe position during unpleasant words, but not pleasant or neutral will confirm our hypothesis. Test for 1c: We will add the factor of Time (early trials, late trials) to the MANOVA in 1a to test this hypothesis. A significant Group[unreadable]Word condition[unreadable]Probe position[unreadable]Time interaction is hypothesized. Post-hoc tests showing BPD>N SEM at 5000 ms during late trials of unpleasant words, but not pleasant or neutral will confirm our hypothesis. Hypothesis (2a): Normals will show greater differential activation (unpleasant-pleasant word condition) in prefrontal regions than in amygdala. Test 2a, prefrontal: For the healthy control group, a repeated-measures MANOVA on BOLD response from the Brodmann program will be conducted. This analysis will allow an examination of prefrontal cortex response to unpleasant vs. pleasant word conditions. The design will be: Word condition (unpleasant, pleasant)[unreadable]Frontal region (dorsolateral, orbitofrontal)[unreadable]Brodmann area (dorsolateral: BA 44, 45, 46;orbitofrontal: BA 11, 12, 47)[unreadable]Hemisphere (right,